Eccentrically Loaded Slip-Resistant Connections

GEOFFREY L. KULAK

ALTHOUGH ECCENTRICALLY loaded connections should be between the bolt a n d the instantaneous center, the
avoided whenever possible, designers are frequendy analysis is quite straightforward.
faced with situations where this type of joint must be used. A single line of bolts loaded eccentrically is shown in
Such cases could be the provision of crane girder brackets Fig. 2. T h e instantaneous center of rotation (i.e.) is lo-
or beam web splices (Figs, l a a n d l b ) . Eccentricity is cated by trial. I t must be on the side of the connection
present even in a standard beam connection (Fig. I c ) , opposite to that of the load and, if the fastener arrange-
although the effect of the eccentricity is usually not sig- ment is symmetrical about a horizontal axis, it will lie on
nificant. that axis. Any trial location of the instantaneous center
Considerable attention has been directed in recent is checked by applying the equations of equilibrium.
years toward the assessment of the ultimate strength of Using the terminology of Fig. 2, these can be stated as
eccentrically loaded connections, whether welded or
fastened with high-strength bolts.^-^'^'^ O n e situation 2 Rs sin 0i = 0 (2)
1= 1
which has received very little attention, however, is the
n
eccentrically loaded bolted connection in which slip is
J2 Rs COS ct>i - P = 0 (3)
undesirable a n d friction-type high-strength bolts are i = i
employed. As far as is known, no tests except those
which will be reported in this presentation have been P(e + ro) - Z r,Rs = 0 (4)
i=i
conducted on this type of connection.^ T h e purpose of
this report is to review the procedure recommended for
the design of slip-resistant eccentrically loaded bolted
connections a n d to compare the results obtained in this
fashion with the available experimental data.

ANALYSIS OF SLIP-RESISTANT CONNECTIONS

Unlike the situation in bearing-type joints, it is reason- it

able to assume that each fastener in a slip-resistant joint 4-^ t
carries the same load. This is true because the slip re-
sistance provided by an individual bolt is the product of
the slip coefficient of the faying surfaces, kg; the number
[b)
of faying surfaces, m; a n d the clamping force provided by
the bolt, Ti\ a n d these quantities are independent of the
location of a given bolt within the joint. Calling the
slip resistance i?s, this can be written as

Rs = mksTi (1)
Considering further that the eccentric connection can be i-f-
assumed to rotate about an instantaneous center and
that the individual bolt forces act perpendicular to a line

Geoffrey L. Kulak is Professor of Civil Engineering, University of (c)

Alberta. Edmonton. Canada.
Fig. 7. Typical eccentrically loaded connections

52

ENGINEERING J0URNAL/AMERICAN I N S T I T U T E OF S T E E L CONSTRUCTION

 shear resistance, even though the friction-type bolt is not
actually acting in shear. At present, the AISG Specifica-
tion permits an equivalent allowable shear stress of 15
ksi for A325 bolts and 20 ksi for A490 bolts.« T h e re-
cently published Guide to Design Criteria for Bolted and
Riveted Joints^ has suggested shear stress values corre-
sponding to various probabilities of slip, a procedure that
tells the designer a good deal more about the potential
I.C.+ ^x behavior of the structure. For example, a slip probability
level of 5 % on A36 steel with clean mill scale corresponds
to equivalent allowable shear stress values for A325 and
A490 bolts of 17.8 ksi and 20.7 ksi, respectively.

DESIGN EXAMPLE
Consider the connection shown in Fig. 3, where five
Fig. 2. Forces on eccentrically loaded slip-resistant connection
%-in. dia. A325 bolts are used in a connection which is
loaded by a force 9 in. from the center of gravity of the
Equations (2) and (3) are more conveniently stated bolt group. T h e bolt resistance will be taken as that
in terms of the x-y coordinate system: corresponding to the equivalent allowable shear stress
level of 17.8 ksi.
(5)
Rs = 17.8 ksi X 0.442 sq. in. X 2 faying surfaces
= 15.7 kips
i?, E - - /• = 0 (6)
i = l ^i
T h e results of the last trial are:
It should be noted that Eq. (2) or (5) will be satisfied
T r y : ro = 1.00 in.
automatically for those cases in which there is no com-
ponent of load in the A:-direction. T h e n : ri = rs = Vv + 6^ = 6.08 in.
T h e trial location of the instantaneous center corre-
r2 = r, = Vv + 32 = 3.16 in.
sponds to the true location when the equations of
equilibrium are satisfied. T h e solution gives both the lo- rs = 1.00 in.
cation of the instantaneous center, which is of no direct
interest to the designer, and the value of the eccentric Solving Eq. (6) for P gives:
load P, which corresponds to the chosen slip resistance .00
per bolt. P = 15.7 yi.oo i.oo\ 2 + 1.00 = 30.9 kips
In order to proceed with a solution, the slip resistance A6.08 3T6/
of an individual bolt, Rs, must be evaluated. Usual prac-
Solving Eq. (4) for P gives:
tice has been to put Eq. (1) into terms of an equivalent
P(9.00 + 1.00) = 15.7[(6.08 + 3.16)2 + 1.00]
P = 30.6 kips
9" Although the values of P so obtained are not identical
< H-^

T ~
i for this trial, they are in close enough agreement.

TEST PROGRAM
Three full-size specimens were tested in order to examine
3/4-0
T the behavior of slip-resistant eccentrically loaded connec-
A325 bolts tions. Considering the variables in the problem, for ex-
I.C.+ 00
ample the condition of the faying surfaces and the varia-
tion in the bolt clamping force, a greater number of
@
specimens would have been desirable. However, these
tests do give a reasonable insight into the behavior of
such connections and allow a comparison between actual
slip loads and predicted values as well as with allowable
Fig. 3. Connection of design example loads.

53
SECOND QUARTER/1975
 100 r
80 h
60 h
LOAD
kips
40 H
20
Fl-4 BOLTS
F2-5 BOLTS
ROTATION , radians x lO'^
Fig. 4. Test specimens and load-rotation behavior
T h e configuration of the test pieces is shown in the
inset to Fig. 4. All specimens used ^^-in. dia. A325 bolts
and all bolts were from the same production lot. Steel
parts used were carbon steel corresponding to A S T M
A36. All faying surfaces were clean mill scale and
standard shop procedures were used in forming the holes,
etc. Bolts were installed in the laboratory, using the
turn-of-ijut method.^ However, actual clamping forces
induced during installation were measured so that slip
loads might be predicted more accurately. No particular
attempt was made to install the bolts in the centers of
their holes and it could be expected that some bolts
might initially be in bearing, just as would be expected
in real structures. T h e average clamping force in these
bolts was about 42 kips, substantially above the specified
minimum load of 28 kips.^ Tests on specimens using
single bolts and loaded axially were also conducted, in
order to determine the slip coeflficient of the faying sur-
faces. It was established that the slip coefliicient of the
material used in the eccentrically loaded specimens was
0.27. Using all of these data, and measured rather than
prescribed dimensions, the loads at which slip should
theoretically have occurred were computed. These are
shown in Table 1.
Neither Specimen 1 nor Specimen 3 showed any
abrupt slip as the connection was loaded. Specimen 1
ENGINEERING J0URNAL/AMERICAN INSTITUTE
Table 1
Slip Loads (kips) Factor of Safety
Speci-
men Theo- AISC ''Guide" AISC "Guide"
retical (Ref. 7) (Ref. 5) (Ref. 7) (Ref. 5)
1 30.0 14.2 19.4 2.11 1.55
2 44.1 20.9 29.3 2.11 1.51
3 63.3 29.4 42.0 2.15 1.51
did show an increased rate of rotation starting at a load
level of about 50 kips. Specimen 2 had an initial slip at a
load level of about 48 kips, compared with the predicted
value of 44.1 kips. Further slips occurred at 53 kips and
63 kips. (The geometry of Specimen 2 corresponds to the
connection used in the Design Example.)
DESIGN RECOMMENDATIONS
At the present time, design of slip-resistant connections
would probably be based on the method prescribed in
the A I S C Manual of Steel Construction J Although that
procedure is fundamentally developed for bearing-type
eccentrically loaded connections, it is based on the use
of a constant resistance for each bolt and is therefore a
satisfactory approach for the slip-resistant case. (In this
method, rotation about the center of gravity is assumed,
however, and the method is not identical to the pro-
cedure described in this paper.)
T h e allowable loads (AISC) are also shown in Table
1. These are calculated on the basis of 15 ksi allowable
equivalent shear stress and the actual rather than an
"eff'ective" eccentricity. T h e introduction of the latter
quantity arose out of tests on bearing-type connections
and is not appropriate for slip-resistant connections. T h e
factor of safety obtained by comparing the theoretical
slip loads with the AISC allowable values is about 2.1
for these specimens.
T h e Guide to Design Criteria for Bolted and Riveted Joints
also recommends allowable loads for eccentrically loaded
slip-resistant connections.^ T h e y use equivalent allow-
able shear stresses corresponding to 5 % or 1 0 % slip
probabilities in combination with coefficients developed
by the writer.^ Allowable loads by this method and
corresponding to the 5 % slip probability level (17.8
ksi for A325 bolts) are also given in Table 1. Comparing
these loads to the theoretical slip loads, it is seen that the
factor of safety is reduced to about 1.5.
SUMMARY AND CONCLUSIONS
Tests of typical slip-resistant eccentrically loaded con-
nections have shown that the actual behavior of the joint
may or may not exhibit slip. If slip does not occur, it is
likely due to the fact that one or more fasteners is already
in bearing or comes into bearing soon after the load is
applied.
54
OF S T E E L CONSTRUCTION
 T h e prediction of allowable loads based on either
current standard practice (AISC Manual of Steel Con-
struction) or recently recommended procedures {Guide to
Design Criteria for Mechanically Fastened Joints) will pro-
duce safe designs. T h e factor of safety provided by the
former method is probably higher than is necessary. T h e
latter method, even at a slip probability method of 5 % ,
produces a design incorporating a reasonable factor of
safety.
ACKNOWLEDGMENTS
T h e work described in this report was conducted with
the assistance of N. M . Holtz and M . J. Perlynn and
their contribution is gratefully acknowledged. T h e study
was carried out in the D e p a r t m e n t of Civil Engineering
at the University of Alberta.
55
REFERENCES
1. Higgins^ T. R. New Formula for Fasteners Loaded Off
Center Engineering News Record, May 21, 1964.
2. Abolitz, A. L. Plastic Design of Eccentrically Loaded Fas-
teners Engineering Journal, AISC, Vol. 3, No. 3, July, 1966.
3. Crawford, S. F. and G. L. Kulak Eccentrically Loaded Bolted
Connections Journal of the Structural Division, ASCE, Vol. 97,
No. ST3, March 1971.
4. Butler, L. J., S. Pal, and G. L. Kulak Eccentrically Loaded
Welded Connections Journal of the Structural Division, ASCE,
Vol. 98, No. ST5, May, 1972.
5. Fisher, J. W., and J. H. A. Struik G*uide to Design Criteria
for Bolted and Riveted Joints John Wiley and Sons, New
York, N. Y., 1974.
6. Specification for the Design, Fabrication and Erection of
Structural Steel for Buildings American Institute of Steel
Construction, New York, N. Y., 1969.
7. Manual of Steel Construction Seventh Edition, American In-
stitute of Steel Construction, New York, N. Y., 1970.
SECOND QUARTER/1975